The proposed studies seek funding to investigate the putative oncogene created in a pre B cell leukemia, characterized by t (1:19) translocation, causing the fusion of the amino terminal segment of transcription factor E2A and a carboxyl terminal segment of PBX1. The resulting protein, when expressed in myeloid cells, causes acute leukemia. Hence, detailed study of the E2A-PBX1 fusion protein is expected to yield insights into both normal and neoplastic cellular development. During the previous tenure of this grant, the Investigators succeeded in identifying a number of functional domains of the E2A-PBX1 oncoprotein. They determined that the fusion protein can form heterodimers with homeo domain containing proteins including several of the HOX genes. A pentapeptide sequence appears to mediate this dimerization. In addition, an amino terminal domain of the E2A fusion partner appears to downmodulate dimer formation, and only binding with the pentapeptide relieves this inhibition. Moreover, a carboxyl terminal domain present in one but not a second spice variant of the fusion protein is also required for heterodimer formation. In using two different assays, transcriptional activation of a reporter gene bearing the specific binding site and a fibroblast transformation assay, two different functions for the fusion protein have been identified. On the basis of these results, a research plan of four specific aims is now proposed. In the first Aim, the Investigator plans to determine how heterodimer formation between the fusion oncoprotein or the normal PBX1 gene and other homeodomain proteins controls DNA binding specificity and transcriptional function. Using a PCR-based assay containing degenerate oligonucleotides, optimal DNA binding sites for PBX1 and six different HOX proteins will be determined. In addition, a non-homeodomain protein, HOX11, will also be evaluated because of its role in the t(10:14) translocational human T cell ALL. The geometry of heterodimer formation will be assessed by comparing the optimal binding sites with the distance between the pentapeptide and homeodomain of each protein. Additional experiments included in this specific aim include the development of a system to determine how the heterodimeric proteins regulate transcription using either yeast or P19 embryocarcinoma cells. In the second Specific Aim, the precise functional sites in heterodimer formation will be mapped. A series of internal mutations will be developed, the resultant proteins expressed and tested for both DNA binding and ability to transform fibroblast in a DNA binding independent fashion. The role of the E2A segment of the fusion protein will be specifically tested by substituting additional transactivation domains from other transcriptional proteins and retesting these two functions. By generating specific mutants of the amino terminus of the E2A-PBX1 protein, the precise site which masks DNA binding by the homeodomain and C terminus of the fusion protein will also be precisely determined. In the third Specific Aim, the Investigators plan to identify additional DNA targets of E2A-PBX1 or PBX proteins using a mobility shift assay. Degenerate oligonucleotides that are flanked by PCR priming sites will be mixed with fusion protein or native PBX1 and then mixed with nuclear extract. Antibody to the PBX portion of the fusion protein or the PBX itself will be used to pull out the complex of DNA and binding partner. PCR will be used to amplify the targeted DNA, after several rounds of selection the bound DNA will be cloned and sequenced. This DNA, in the presence of a fusion protein or PBX will then be used to clone the fusion partners using a lambda GT11 expression system. Finally, in the fourth specific aim, the role of PBX1 will be determined by generating homologous recombinants using the Cre- lox recombination system. The Investigators postulate that knock- out of PBX1 will likely be lethal, given its widespread expression. Hence, a newly developed system by which knock-out can be directly targeted in a tissue-specific or developmental stage specific manner will be employed. By incorporating three recombination sites flanking the homeodomain of PBX1 gene in homologous recombinants, back crosses to animals in which the recombinase is expressed in a tissue or developmental specific manner will allow elimination of PBX1 only in those tissues or only at that time during development that the promoter driving the recombinase is active. Once obtained, RNA will be obtained from the targeted tissues to confirm specific elimination of PBX1 function. In this way, using representational display analysis-PCR specific genes which are up or downmodulated by PBX1 will be isolated and ultimately cloned.